The purpose of these studies is to define the types and mechanisms of action of T cells which regulate antibody responses to the so-called thymus-independent (TI) antigens type III pneumococcal polysaccharide (S3) and polyvinylpyrrolidone (PVP). Two distinct subsets of S3-specific suppressor T cells (T[unreadable]S[unreadable]) have been characterized, and these T[unreadable]S[unreadable] suppress the anti-S3 antibody response by distinct mechanisms. One of the S3-specific T[unreadable]S[unreadable] has apparent specificity for S3 plus self I-J determinants. These T[unreadable]S[unreadable] are apparently induced by S3 presented on a subset of antigen presenting cells that are cyclophosphamide (Cy)-sensitive and bear I-J determinants. Recent studies have also established that a contrasuppressor T cell (Tcs) is induced by S3 or by S3-coupled spleen cells (S3-SC). This T-cell subset interferes with the activity of S3-specific T[unreadable]S[unreadable] so that the latter cells can only be detected in situations where Tcs activity is deleted, e.g., by treatment of mice with Cy or by removal of Tcs on V. villosa lectin. The mechanism by which Tcs regulate T[unreadable]S[unreadable] activity is currently being investigated. Tcs are activated only by immunogenic doses of S3 and are not present in mice tolerant to S3. These and other results suggest that S3-specific B cells may be involved in Tcs activation. In the PVP system, two subsets of T[unreadable]S[unreadable] have also been defined, and these have several different characteristics from the T[unreadable]S[unreadable] induced by S3. We have also shown that very low (subimmunogenic) doses of PVP can activate specific helper T cells (T[unreadable]H[unreadable]), while doses of PVP that induce optimal antibody responses activate both T[unreadable]H[unreadable] and T[unreadable]S[unreadable], although the latter predominate so that T[unreadable]H[unreadable] can be detected only when T[unreadable]S[unreadable] are eliminated. Preliminary studies suggest that this is also true for S3. This suggests that if T[unreadable]S[unreadable] activation can be reduced (or eliminated) some of the so-called TI antigens can behave similarly to T-dependent antigens. Since many of the bacterial pathogens, such as pneumococci and meningococci, elicit primarily TI-type antibody responses, it may be possible to improve the efficacy of bacterial vaccines by immunizing under conditions that would favor T[unreadable]H[unreadable] rather than T[unreadable]S[unreadable] activation. Our basic studies may ultimately prove to be applicable towards achieving such a goal. (LB)